Importing third party packages crashes kernel of Jupyter notebooks in VScode #25700
Description
Environment data
- VS Code version: 1.107.0
- Jupyter Extension version (available under the Extensions sidebar): 2025.9.1
- Python Extension version (available under the Extensions sidebar): 2025.20.0
- OS (Windows | Mac | Linux distro) and version: Windows 11 Enterprise (Version 23H2)
- Python and/or Anaconda version: Miniforge_BASE_24.11.3; conda 24.11.3; mamba 1.5.12
- Type of virtual environment used (N/A | venv | virtualenv | conda | ...): mamba
- Jupyter server running: Local
Expected behaviour
Until this morning, I could use my Python environments and create new ones with e.g. mamba create -n <name> -c conda-forge python=3.12 numpy jupyter ... which would work in VScode and I could import any installed package.
Actual behaviour
Now, whenever I try to import a third party package, the kernel dies (even for freshly installed environments) and the logs show the following errors:
[error] Failed to get activated conda env vars for c:\Trusted\test_env2\python.exe
[error] Unable to determine site packages path for python c:\Trusted\test_env2\python.exe (Conda)
The kernel nevertheless starts successfully and can import packages like os and generally execute cells. Only third party packages cause the kernel crash. When instead activating an environment in the command line and starting jupyter from there, everything works normally. Also, my old environments still work as intended, as well, even in VScode.
Noteworthy, the Python environments are stored in this Trusted folder while mamba is installed under C:\Program Files\miniforge3\Scripts\mamba.exe as per company policy. However, until now this setup worked.
Steps to reproduce:
If the installation paths are the culprit, I am not sure how this would be easily reproducible elsewhere.
Logs
